slisemap - Version 1.6.2

Added a reference to the published SLIPMAP paper.

- Python
Published by speug almost 2 years ago

slisemap - Version 1.6.1

What's Changed

• Faster kmeans when plotting: https://github.com/edahelsinki/slisemap/pull/15

- Python
Published by Aggrathon about 2 years ago

slisemap - Version 1.6.0

What's Changed

• Implement Slipmap, a variant based on prototypes that scales better with respect to both time and memory.
• Separate (optional) regularisation functions.
• Use Bayesian optimisation for hyperparameter tuning.
• Remove pre-1.3 deprecations.
• Add optimisation stats to metadata.
• Move paper experiments to frozen branches

Full Changelog: https://github.com/edahelsinki/slisemap/compare/v1.5.2...v1.6.0

- Python
Published by Aggrathon about 2 years ago

slisemap - Version 1.5.2

Changes:

• Improved sm.plot_dist(scatter=True)
• Improved cluster handling in sm.plot()
• Added example notebook for interactive plots

- Python
Published by Aggrathon almost 3 years ago

slisemap - Version 1.5.1

Changes:

• Fix a isinstance(bool, int) == True oversight

- Python
Published by Aggrathon almost 3 years ago

slisemap - Version 1.5.0

Changes: - plot: improve labels for matrix plot - plot: more flexible clusters (not just integers [0..k]) - plot: bars accepts a list of variables - get_model_clusters: sort the clusters labels based on the embedding - save and load: add (optional but default) LZMA compression - metadata.get_variables: fix fallback

- Python
Published by Aggrathon almost 3 years ago

slisemap - Version 1.4.0

Additions: - Add procedures for hyperparameter optimisation (see slisemap.tuning). - Added wrappers that combine local_model, local_loss, and coefficients into one name (see, e.g., slisemap.local_models.LogisticRegression).

Changes: - Change defaults for slisemap.metrics.accuracy (to optimise=False, local_loss=True). - Add numpy parameter to Slisemap.fit_new. - Improve the Slisemap.predict function to take more parameters. - Move entropy to slisemap.metrics. - Fix some deficiencies in slisemap.utils.Metadata. - Improve the x-axis labels for the matrix plot.

- Python
Published by Aggrathon about 3 years ago

slisemap - Version 1.3.1

New features:

• Metadata
  • Add metadata about e.g. variable names and normalisation
  • The metadata is reused for all plots
• More general input
  • Anything that can be turned into a torch.Tensor or numpy.ndarray is accepted
  • This includes dataframe-like objects (also imports variable names)

Improvements:

• Better colourscales for local losses
• Centering before PCA
• Slisemap.load understands map_location from torch.load
• More verbosity levels for optimisation

Changes:

• Remove types from docstrings
• Rename "Fidelity" to "Local loss" in plots

Deprecations:

• With the addition of metadata, some parameters to the plotting functions have been deprecated

- Python
Published by Aggrathon about 3 years ago

slisemap - Version 1.2.1

What's Changed

• Refactor demo examples + add binder demo to README by @MomoLangenstein in https://github.com/edahelsinki/slisemap/pull/3
• Point the user towards multiplelinearregression if they use nd-y by @MomoLangenstein in https://github.com/edahelsinki/slisemap/pull/5
• Improve docs by @Aggrathon in https://github.com/edahelsinki/slisemap/pull/6
• Cleanup of properties (some deprecation warnings)
• Update references by @Aggrathon in https://github.com/edahelsinki/slisemap/pull/7

Full Changelog: https://github.com/edahelsinki/slisemap/compare/v1.1.0...v1.2.1

- Python
Published by Aggrathon about 3 years ago

slisemap - Version 1.1.0

Main changes in this release:

• Improved documentation.
• Less randomness (only perturb the embedding if a loss of rank is detected).
• New argument only_B in the Slisemap.optimise (the same as Slisemap.lbfgs(only_B)).
• Tweaks to the plots and updated experiments.
• A notebook discussing how to use PyTorch for optimisation.

- Python
Published by Aggrathon over 3 years ago

slisemap - Version 1.0.4

• Exposed get_model_clusters.
• Made scikit-learn a required dependency (and removed all optional dependencies).
• Fixed a bug when calculating PCA using SVD.

- Python
Published by Aggrathon over 3 years ago

slisemap - Version 1.0.2

• Fix a bug with the scaling in fit_new
• More verbosity in fit_new and optimise
• Add logistic regression with log-loss to local_models

- Python
Published by Aggrathon almost 4 years ago

slisemap - Version 1.0.1

• Fix a bug related to saving when using random_state
• Improve the documentation

Attached to this release is the cached results from running the experiments in the paper (for easier reproducibility).

- Python
Published by Aggrathon almost 4 years ago

slisemap - Version 1.0

The package now contains all the functionality mention in the paper and the demo paper. Additionally, the experiments directory contains the code for all the experiments mention in the paper and the examples directory contains the notebook mentioned in the demo paper. Attached to this release is the cached results from running the experiments in the paper (for easier reproducibility).

- Python
Published by Aggrathon almost 4 years ago

slisemap - Documentation

• Use sphinx to generate a documentation website
• Add a random_state parameter
• Use __slots__ in the Slisemap class for speed improvement and typo checking

- Python
Published by Aggrathon almost 4 years ago